cs.CV（2024-07-10）

📊 共 18 篇论文 | 🔗 9 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (7 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗4) 支柱三：空间感知与语义 (Perception & Semantics) (3 🔗2) 支柱一：机器人控制 (Robot Control) (2) 支柱六：视频提取与匹配 (Video Extraction) (2 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#	题目	一句话要点	标签	🔗	⭐
1	LEMoN: Label Error Detection using Multimodal Neighbors	LEMoN：利用多模态邻域检测图像-文本数据集中错误标注	multimodal
2	LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models	LLaVA-NeXT-Interleave：统一处理多图、视频和3D场景的大型多模态模型	multimodal	✅
3	RoBus: A Multimodal Dataset for Controllable Road Networks and Building Layouts Generation	RoBus：用于可控道路网络和建筑布局生成的多模态数据集	multimodal	✅
4	MAN TruckScenes: A multimodal dataset for autonomous trucking in diverse conditions	发布MAN TruckScenes多模态数据集，助力自动驾驶卡车在复杂环境下的感知研究。	multimodal
5	Controlling Space and Time with Diffusion Models	提出4DiM，一种用于自然场景4D新视角合成的级联扩散模型	TAMP
6	InstructLayout: Instruction-Driven 2D and 3D Layout Synthesis with Semantic Graph Prior	InstructLayout：提出一种结合语义图先验的指令驱动2D/3D布局合成框架。	multimodal
7	Pseudo-RIS: Distinctive Pseudo-supervision Generation for Referring Image Segmentation	提出伪RIS框架以解决图像分割中的伪监督生成问题	foundation model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#	题目	一句话要点	标签	🔗	⭐
8	TIP: Tabular-Image Pre-training for Multimodal Classification with Incomplete Data	提出TIP框架，解决不完整表格数据下表格-图像多模态分类问题	representation learning contrastive learning multimodal	✅
9	Protecting NeRFs' Copyright via Plug-And-Play Watermarking Base Model	提出NeRFProtector以保护NeRF版权问题	distillation NeRF neural radiance field	✅
10	MambaVision: A Hybrid Mamba-Transformer Vision Backbone	MambaVision：一种混合Mamba-Transformer视觉骨干网络，提升视觉特征建模效率。	Mamba	✅
11	Simplifying Source-Free Domain Adaptation for Object Detection: Effective Self-Training Strategies and Performance Insights	提出简单有效的自训练策略，显著提升无源域目标检测的性能。	teacher-student IMoS	✅

🔬 支柱三：空间感知与语义 (Perception & Semantics) (3 篇)

#	题目	一句话要点	标签	🔗	⭐
12	OV-DINO: Unified Open-Vocabulary Detection with Language-Aware Selective Fusion	OV-DINO：提出语言感知选择性融合的统一开放词汇目标检测框架	open-vocabulary open vocabulary	✅
13	Flow4D: Leveraging 4D Voxel Network for LiDAR Scene Flow Estimation	提出Flow4D以解决LiDAR场景流估计中的时空特征提取问题	scene flow	✅
14	Swiss DINO: Efficient and Versatile Vision Framework for On-device Personal Object Search	Swiss DINO：高效且通用的设备端个人物体搜索视觉框架	scene understanding

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#	题目	一句话要点	标签	🔗	⭐
15	Controllable Navigation Instruction Generation with Chain of Thought Prompting	提出C-Instructor，利用思维链提示实现可控的导航指令生成	manipulation large language model chain-of-thought
16	Disentangling Masked Autoencoders for Unsupervised Domain Generalization	提出DisMAE，通过解耦表征学习实现无监督领域泛化	manipulation masked autoencoder distillation

🔬 支柱六：视频提取与匹配 (Video Extraction) (2 篇)

#	题目	一句话要点	标签	🔗	⭐
17	Hybrid Structure-from-Motion and Camera Relocalization for Enhanced Egocentric Localization	提出混合结构光重建与相机重定位以增强自我中心定位	egocentric	✅
18	Raising the Ceiling: Conflict-Free Local Feature Matching with Dynamic View Switching	提出RCM：通过动态视角切换实现无冲突的局部特征匹配，显著提升匹配数量。	feature matching

⬅️ 返回 cs.CV 首页 · 🏠 返回主页